How MRT Rail Lubrication and Friction Management Work Using Mathematics

A railway needs friction to accelerate and brake. The same friction can also create wear, lateral force, corrugation and curve squeal.

Friction management is therefore not the removal of friction. It is the controlled shaping of friction at different parts of the wheel–rail interface so traction remains usable while unnecessary wear and instability are reduced.

The canonical Wheel–Rail Contact pillar explains how traction, braking, creepage and guidance arise.

The Grinding and Wheel Profiling pillar owns geometry restoration.

The Noise and Vibration pillar owns the received acoustic and vibration field.

This article owns a narrower layer:

How can a railway deliberately modify wheel–rail friction in selected contact regions so the interface remains sufficiently adhesive for safe traction and braking while avoiding more friction, wear or instability than the job requires?

LTA has publicly identified curved track as a source of higher wheel–rail noise on parts of the Thomson-East Coast Line and described speed reduction and rail grinding as mitigation. More broadly, railway research has shown that top-of-rail friction modifiers can reduce curve squeal and alter corrugation growth by controlling the friction–creep relationship. This article uses that general engineering literature without claiming a particular friction-modifier product or application programme is currently used on Singapore’s MRT.

The RFE — What Is Friction Management For?

The weak objective is:

make the rail slippery

That would be dangerous and conceptually wrong.

The Reason for Existence is to keep friction inside the right functional range at the right contact location so useful adhesion survives while excessive wear, lateral force, noise and unstable stick–slip behaviour are reduced.

Prompt 1 — Why Is More Friction Not Always Better?

Available tangential force is bounded approximately by:

|Ft| ≤ μN

A larger μ can increase available adhesion.

But in curves it can also increase lateral creep force and wear.

Frictional energy dissipated in a sliding or creepage process scales as:

Pfriction = Ft vslip

More tangential force at the same slip velocity means more frictional power entering heat, wear and vibration.

The useful target is therefore not maximum μ.

It is sufficient and stable μ.

Prompt 2 — Why Are the Gauge Face and Top of Rail Different Jobs?

The wheel flange can contact the rail gauge face strongly in curves.

Traditional gauge-face lubrication aims mainly to reduce flange/gauge-face friction and wear.

The top of rail is different.

The wheel tread on the rail head must still transmit traction and braking.

gauge face:
reduce unnecessary sliding friction and wear

top of rail:
control friction without destroying useful adhesion

This is why top-of-rail materials are often called friction modifiers rather than simple lubricants.

Prompt 3 — What Is a Friction–Creep Curve?

At small creepage ξ, tangential force often rises approximately linearly:

Ft ≈ Cξ ξ

At larger creepage, the contact approaches saturation:

|Ft| → μN

Research on curve squeal shows why the slope beyond saturation matters. If increasing slip is associated with falling friction force, the contact can support unstable roll–slip oscillation.

A friction modifier aims to shape the relationship towards a controlled intermediate friction level and more stable positive friction behaviour.

The details depend on material and interface state; this article does not prescribe products or coefficients.

Prompt 4 — How Can Friction Create Curve Squeal?

A wheelset negotiating a curve develops lateral creepage and contact forces.

If frictional feedback becomes unstable, the wheel can alternate between higher and lower slip states and excite a structural resonance.

lateral creepage
→ friction instability
→ oscillatory tangential force
→ wheel vibration mode
→ radiated squeal

A simplified self-excitation condition can be thought of as negative effective damping:

ceffective = cstructure + cfriction

If friction contributes enough negative damping that c_effective<0, oscillation can grow.

Friction management attempts to change that feedback, while the noise pillar owns the acoustic result.

Prompt 5 — How Does Friction Affect Wear?

A common first model is Archard wear:

Vwear = K Ns/H

But wheel–rail wear depends strongly on contact pressure, creepage, material state, contamination and frictional energy.

An energy-based wear index can be written conceptually:

Windex = ∫ |Ft vslip| dt

Reducing unnecessary tangential force or slip can therefore reduce wear energy.

Yet reducing friction too far risks weakening adhesion.

Friction management is optimisation under a hard floor: wear can be reduced only while traction and braking remain adequately supported.

Prompt 6 — How Much Material Should Be Applied?

This is precisely where a public article should stop before becoming an operating recipe.

At a conceptual level, let application rate be q and achieved interface state be μ(q).

The optimisation is:

choose q to minimise
wear + squeal + lateral-force penalty + material cost

subject to
traction/braking adhesion remaining acceptable

Real application quantities, locations, products and operational limits depend on authorised engineering validation and are intentionally not given here.

Prompt 7 — How Does Weather Change the Interface?

Wheel–rail friction is affected by contaminants and environmental conditions.

A simplified state might be:

μ = μ(material film, water, contamination, temperature, traffic history)

A friction-management system therefore needs robustness across changing real-world states rather than performance in one clean laboratory condition.

This is another reason measured World Return matters.

Prompt 8 — How Does the Railway Know Friction Management Worked?

Possible outcome metrics include:

curve noise level;
wheel or rail wear rate;
corrugation growth;
lateral force;
traction or braking performance;
material consumption;
maintenance interval.

For before/after wear rate w:

Δw = wafter − wbefore

For noise level L:

ΔL = Lafter − Lbefore

But a quieter curve alone is not sufficient proof.

The railway must also verify that adhesion and braking performance remain within their authorised envelope.

A Fictional Friction-Management Example

Consider a fictional curve where one contact patch carries normal load N=70 kN.

Dry effective friction coefficient is μ=0.45.

Fmax≈μN
    ≈31.5 kN

Suppose a fictional friction-management state changes effective μ to 0.32 while remaining inside the validated adhesion range.

Fmax≈22.4 kN

If the curve needed only 15 kN of tangential force for the relevant guidance state, both friction states provide enough theoretical margin, but the lower controlled state may reduce excess frictional energy.

At slip velocity 0.20 m/s:

Pfriction,dry=31.5×0.20≈6.3 kW
Pfriction,controlled=22.4×0.20≈4.5 kW

This is only an illustration; real wheel–rail force is governed by creepage and saturation rather than simply sitting at μN.

Deletion Tests and Failure Shadows

Remove adhesion: friction can be reduced without affecting traction or braking.
Remove gauge-face/top-of-rail distinction: every contact region has the same friction job.
Remove creepage: tangential force is independent of relative rolling motion.
Remove friction instability: curve squeal cannot originate from the interface.
Remove wear: excessive friction has no lifecycle cost.
Remove environment: water and contamination never change the interface.
Remove application uncertainty: any quantity of modifier produces exactly the desired friction.
Remove World Return: quieter operation is accepted without checking adhesion or wear.

The Friction-Management Audit

Which contact region is being managed?
What function must friction preserve there?
What traction or braking floor must remain?
What creepage and lateral-force state exists?
What wear, corrugation or noise problem is being targeted?
What environmental state changes friction?
How will material application be controlled and validated?
What before/after measurements prove benefit?
What evidence would show friction was reduced too far?
How do grinding, profiling and contact geometry interact with the friction intervention?

World Return — The Wheel Decides Whether the Film Worked

measure baseline contact consequences
→ apply authorised friction-management intervention
→ trains pass
→ measure noise, wear, forces and adhesion proxies
→ compare
→ adjust or remove intervention
→ continue monitoring

MRT friction management works when the wheel–rail interface retains enough friction to do the railway job while giving up the excess friction that mainly creates wear, noise and instability.

Reader-safety note: This article does not prescribe friction-modifier products, application rates, rail locations, adhesion thresholds or operating procedures. It does not claim that a particular top-of-rail friction-modifier system is currently deployed on Singapore’s MRT unless an operator or LTA publicly states so.

How MRT Rail Lubrication and Friction Management Work Using Mathematics: Why the Best Wheel–Rail Friction Is Not Always the Most